Feeds

How I learned to stop worrying, and abandoned Mac OSX

New Coke, meet Classic

  • alert
  • submit to reddit

Abandoned Christmas trees are piling up on the sidewalks in the Silicon Valley suburbs, but I'm hoping that some of the season's goodwill hasn't been discarded with them. In fact, I'm counting on it - for I'm breaking both a personal and a Register house rule with the following story. It contains stuff you already know, or stuff you suspected was true, or stuff you have already found more comprehensively and eloquently expressed elsewhere.

As a rule, we don't waste your time with solipsistic rants unless they impart some new information or ideas into the public realm [so get on with it - ed.].

Both John Siracusa at Ars Technica and Bruce Tognazzini have raised the same concerns, with Tog warning that Apple's dismal OS X user interface was leading the company into a New-Coke style disaster. But if we can indulge you, this is a battle-tested road report on rubbing along with OS X. That's eight months spent on our own personal kit, trying to justify the investment. And watching the gold CD-ROM cursor spinning, and spinning.

It's the UI, Stupid

Why should the UI matter? And why does it matter so particularly to Apple loyalists?

Well, satellite guru and erstwhile Kaleidoscope scheme author Lloyd Wood summed it up pretty sweetly for us. (You might have recall Wood from his celebrated demolition of Jon Katz).

The Mac had become nothing but a UI, he said, justifying his decision to abandon the platform. Behind the perfect UI, what technology is there hangs together thanks to string and chewing gum. And he's right. The technology behind MacOS is truly awful and should have been put out to pasture by the early nineties. What other modern OS goes into a coma when you linger over a menu for a moment too long? And Mac users have suffered unnecessary crashes, truncated file names (31 characters really isn't such a big deal anymore) far longer than they needed to. Much of that has been cured in OS X, but evidently all wasn't well.

However, I figured that given the human capacity to adjust to new environments, I'd slide into the new UI without too many bruises. This didn't happen, and a few weeks ago reader Tim Priest who set me thinking the unthinkable with these thoughts:

"OS X has some good features. At this time user interface is not one of them," he wrote a month ago.
"Unfortunately, this is the one thing that has kept Apple buoyant for so long. I know Mac users that are now considering ditching the Mac based solely on the horrid user experience and usability issues that OS X is foisting upon them. I may well be one too."

Which pretty much summed up how my experience. Then, shortly before Christmas, genial pragmatist Henry Norr confessed in the Chron: "After months of forcing myself to use Mac OS X, I've gone back to Mac OS 9 and Office 2001."

Death by a thousand cuts

And so did I, five weeks ago. With one simple click of the startup folder preference panel, we were home again.

I made a pretty simple calculation. I hadn't abandoned OS X because I was disappointed with a functionality deficit - it does a everything a modern OS should do. And then some - the near instant power-on/sleep of the Mac has been improved, and is unique in a modern desktop PC. And it has the best of the old and the new: Cocoa services and AppleScript are powerful environments, which again Apple should be proud of as unique to the platform.

No it's simple: it's the User Interface, Stupid. We'll be particularly cruel here and bracket performance in as a user interface issue. It's really what marketing types would call a "user experience" issue, but if performance impairs the usage of the machine, then that's bad UI. And performance on OS X is really NOT GOOD.

Nor is it the file typing and meta data issue that's been so vexing users.

It's been the daily "stickiness" that's been so infuriating: Aqua has been more a case of a death of a thousand cuts: there's no big flaw so obnoxious it's enough to cause a revolt , but it's felt like a constant hindrance, and the annoyances add up. Here are a few.

I've got a little list

1. It's Sticky: Although Spring Loaded folders and pop-up windows were relatively recent additions to the Mac UI, they're great examples of flexibility that disappeared in Aqua. If you need to move a file from a desktop into say a 'Work' folder in your Documents 'folder', you're asked to complete a three stage operation. Given the window clutter that might be in the way, that's probably reaching double figures.

But there's more to 'stickiness' than feature withdrawal. The best analogy for Aqua is of a map of the world in which individual cartographers were each drawing a continent, but working off different co-ordinate systems. There's little feel for a cohesive whole, or time-tested user labs over-ruling some of the absurdities.

For example, where does the Dock go? You'll soon find the practicalities overrule aesthetics: for two of the three locations cause niggling problems.

Unless it's been hidden manually, the Dock is always visible, and guaranteed to overlay the foreground application window. Now, many Carbon applications don't know where the Dock is, so if you maximise an application, the window's resize handle (and there's only one) is now inaccessible. So off you go, and hide the dock, just so you can grab the handle and resize the window?

No problem, you say, just put the Dock on the Left, where it looks kinda neat (and where, I notice, the Apple helpers in the San Francisco branch of CompUSA like to put it at the start of the day on the store's demo machines). But then you get another example of the mismatched UI klutziness: new window instances don't know where the Dock is either. And the Aqua geniuses have put the three control buttons (minimize, maximize, close) on the left hand side. So a new Finder window will appear with the windows controls obscured. That means giving each new window a little rightward shuffle. It's a tiny niggle, but amplified by repetition. My Dock's on the right, which means the arrows point right when the menu springs left, which is plain ugly. But it's the only way to survive.

Stickiness is manifest in other ways, too. A bugbear of mine is the watery-blue sliders in the scroll bars. They feel far more elusive than the utilitarian blocks of grey in the old MacOS, and every other OS.

2. It's Slow:I dare you. Reboot into classic, fire up IE and feel the difference. Despite the speed improvements in 10.1, Aqua behaves far too slowly. Scrolling in the Finder feels still feels jerky, even though the raw I/O performance of X makes far better use of the hardware than clunky Classic. (Try copying a well-populated folder).

But even throwing a 733Mhz G4, with half a GB of memory, at Aqua has worrying side effects. The sheer horsepower of producing some of the Aqua Eye Candy has caused me skips when replaying MP3 files. And I've experienced peculiar, unwanted cut-up effects in the Carbon Sound Studio which may be bugs, but I half suspect are serious timing issues introduced by the CPU cycle-sapping UI.

3. It's Patronising: Face it, you're using one of those large-print books created for adults with reading difficulties. Very useful for people with serious learning difficulties, of course, but immensely patronising for the rest of us, and few OS X reviewers - and John Siracusa is a noble exception - have brought up how big and dumb the new UI looks.

With OS X, it feels very much like your real estate has evaporated. I ran classic in 1024x768 (for the higher refresh rate), and the endless steppes of a 1096 pixel resolution practically gave me agoraphobia.
But in X, I need to crank up to 1280x1024 to get the same work done.

In truth it's down to perception as much as reality: the menu bar takes up no more space in X than in the vintage MacOS. However the default icon font remains absurdly large, and the solution to that has already been crushed by the next great drawback of OS X:-

4. Obey The Steve: Which is where the free-thinking, free-wheeling image Apple has promoted really collapses. Apple has decided that the individualists who use the software can't change a thing about it. You can't even decide not to display no desktop wallpaper, let alone replace the default shell. By contrast, in the wicked world of Windows, shell replacements (or enhancements) like Litestep or DesktopX. Some gaps have been plugged: ASM provides a pull down switcher, but I badly miss LiteSwitch, the keyboard switcher than also let you kill and hide apps as you Alt-Tabbed your way through the task list.

The Dock supports task switching from the keyboard, but incredibly, only in the order that applications appear in the dock. Which is next to useless.
Were Apple to document Quartz APIs many of these grumbles would go away. Kaleidoscope schemes could make a big difference here, for example. As it is, it's adding insult to injury. Apple is laying down quite a cynical challenge to its most loyal users: either trust us and go with us, or beetle off.

5. When Metaphors Roamed The Earth: Docks aren't necessarily evil. The popularity of the NeXT-alike shell WinSTEP on Windows, and the WindowMaker window manager on Linux shows that NeXT's original UI endures. It's a nice way to organise things.

But OS X's Dock is a truly terrible Dock, solely because of Apple's decision to base it for essentially incompatible UI tasks: launching, switching.

6. The Dock Must Go:
By insisting on some traits of NeXT-ish behaviour, but bowing to the most vociferous Apple loyalists, OS X can at times be a horrible hybrid, and you can reasonably argue that Apple has ruined not just one great UI, but two. NeXT veterans who know they're essentially using a port of the OS rue the loss of unique NeXT features, principally services.

(You can get a glimpse of how great NeXT services can be in the Stickies app - it can send text to a browser, email or rich text document with one click). But by basing the OS X shell on Carbon's immature APIs (it's only just become services-aware) Apple was presenting the worst face it could to the world. Which brings us to where we are now.

Budget cuts home to roost

What Apple can do and what Apple should do are two different questions. You can probably guess that I'm not as enamoured of the spatial desktop as some UI purists.

This is an office automation metaphor hangover from the 1970s that works great for folk using disconnected computers with a few dozens or at most hundreds of files. And it works when applications store stuff in discrete files, rather than their own data silo. For example your email client of choice almost certainly stores data in such a format that leaves the application itself as the only reasonable viewer and retrieval mechanism. (And, yes, Eudora stores its mailboxes in plain text format, but Sherlock isn't much use here).

I'd be happy to give up the Mac UI for something better, a UI that lets me tag and categorise information on the fly. We need ways of caching information, just long enough to personalise it.
Microsoft is promising to put a database at the heart of future versions of Windows (it may be SQL Server, it may not) and run the file system as a legacy device driver, which should help with data retrieval enormously. (As well as give Microsoft the default file format for every application for all time).

But Apple abandoned hope of matching such bold initiatives when Jobs axed the R&D budget, despite many man years of investment, when he resumed the leadership. One of the first casualties was OpenDoc, which wouldn't have solved the information overload problem of a Net-attached computer, but would have at least have laid the ground for new UIs ideas.

Apple doesn't need to turn the tables in quite such a fashion as Windows Blackcomb promises: BeOS didn't use a database as the native file storage but the rich file system attributes of BFS gave it many of the same properties. However BeOS was only capable of such spectacular real-time queries because the OS latencies were so low: BeOS journalled the file attributes on the fly, and unless you have a speedy OS, you'll have the crash and grind familiar to Sherlock or Fast Find users.

The best we can hope for is that the successor to HFS+, the Mac's current file system, includes many of these features. And that in the interim, the existing Finder/Dock combination is superseded by something more speedy, practical and graceful. ®

Bootnote:
Any suggestions for what to do with my G4? I'll give OS X another whirl with 10.2. It's by far my noisiest computer, and so makes the least lovely candidate for serving or routing duties. There's Linux of course, which I know well. But Linux on non-x86 platforms lacks the applications you expect like native Real, Flash and Yahoo IM clients, for example. It could be fun watching X and Gnome mature on Darwin.

There are still other reasons to covet Apple computers - the helpful community, the above average build quality (we'll charitably overlook the recent run of iffy mice and keyboards), and the buoyant after-market which means your investment is still worth something three years down the line - well you'll have your own favourites. But the Apple user interface is no longer one of them. In fact we have to conclude that with OS X, you're buying Apple in spite of the user interface, not because of it.

Whitepapers

Designing a defence for mobile apps
In this whitepaper learn the various considerations for defending mobile applications; from the mobile application architecture itself to the myriad testing technologies needed to properly assess mobile applications risk.
3 Big data security analytics techniques
Applying these Big Data security analytics techniques can help you make your business safer by detecting attacks early, before significant damage is done.
Five 3D headsets to be won!
We were so impressed by the Durovis Dive headset we’ve asked the company to give some away to Reg readers.
The benefits of software based PBX
Why you should break free from your proprietary PBX and how to leverage your existing server hardware.
Securing web applications made simple and scalable
In this whitepaper learn how automated security testing can provide a simple and scalable way to protect your web applications.